GaggiX

42 ms·

721.
by GaggiX 2y ago
>Attention Is All You Need paper detailing transformers, and it would be an entirely different market if Google had held that as a trade secret. I would guess that in that timeline, Google would never have been able to learn about the in
722.
by GaggiX 2y ago
>For example, has anyone ever attempted image -> html/css model? Have you tried upload the image to a LLM with vision capabilities like GPT-4o or Claude 3.5 Sonnet?
723.
by GaggiX 2y ago
Llama 3.1 405B is on par with GPT-4o and Claude 3.5 Sonnet, the 70B model is better than GPT 3.5 turbo, incredible.
724.
Azure Llama 3.1 Benchmarks(old.reddit.com)
4 points by GaggiX 2y ago|0 comments
725.
by GaggiX 2y ago
These ones are quite easy, the most difficult ones are the ones created with MJ with style raw and prompts like "snapchat photo".
726.
by GaggiX 2y ago
>Those restrictions are set up through the Robots Exclusion Protocol Well so it's not really disappearing at all.
727.
by GaggiX 2y ago
Multiplication requires O(n^2) complexity with the usual algorithm used by humans, LLMs have a constant amount of computation available and they are not really efficient machines for math evaluation. They can definitely evaluate unseen expr
728.
by GaggiX 2y ago
GPT-3 was 173B parameters and it's very bad compare to much smaller models we have nowadays, the data and the compute play a giant role, also I doubt you would need to train a model further after you have trained it on absolute everyth
729.
by GaggiX 2y ago
It's clearly working because the models are only getting better, believing that the performance of these models would fall at some point in the future is just very delusional.
730.
by GaggiX 2y ago
The post was probably written by a mere human (they do sometimes hallucinate quite badly).
731.
by GaggiX 2y ago
Well that's even better for the point of the article.
732.
by GaggiX 2y ago
>If BERT worked so well, why not scale it? I mean, the scaling already happened in 2019 with RoBERTa, my guess is that these models are already good enough at what they need to do (creating meaningful text embeddings), and making them ex
733.
by GaggiX 2y ago
Did they try using this method on something that is not StyleGAN?
734.
by GaggiX 2y ago
This is not really how these models work, if the sample is out of distribution then it would usually perform worse on the task assigned.
735.
by GaggiX 2y ago
>AI is one of the few technologies that doesn't just fail to do it's job but it fails and you might never notice until the error is already very costly if it hallucinated something crazy. Because this is what is used to deal wi
736.
by GaggiX 2y ago
>In pre-training, we filter out(opens in a new window) information that we do not want our models to learn from or output, such as hate speech, adult content, sites that primarily aggregate personal information, and spam. Great so now th
737.
Solid State Batteries Are Here: Yoshino Power Station(undecidedmf.com)
3 points by GaggiX 2y ago|0 comments
738.
by GaggiX 2y ago
This is about the Pile dataset, of course we don't if it has been used to train the commercial models we use or just for the research papers mentioned in the article.
739.
by GaggiX 2y ago
3.5 Sonnet outperforms GPT-4o on most benchmarks.
740.
by GaggiX 2y ago
Are humans provably correct and ethical in their outputs?
741.
by GaggiX 2y ago
The feature was enable by the author.
742.
by GaggiX 2y ago
What would you use it for, a kid?
743.
by GaggiX 2y ago
I missed that, thank you.
744.
by GaggiX 2y ago
The perfect smartphone to smuggle.
745.
by GaggiX 2y ago
I asked for the most densely populated square kilometer, not the most densely populated city. A city's population can be spread out evenly or concentrated in a few areas.
746.
by GaggiX 2y ago
I don't think there is a browser extension but there is a site called websim AI that creates fake pages in real-time using LLMs and it honestly works surprisingly well.
747.
by GaggiX 2y ago
Now I wonder what would be the most densely populated square kilometer in the world instead. I would imagine something like in Hong Kong.
748.
by GaggiX 2y ago
What AIs are you referring to? "If ... Else" type of AI?
749.
by GaggiX 2y ago
This is incredibly delusional ahaha Just look at how popular TikTok is (it wouldn't be without its powerful recommendation model), or the fact that the majority of the world's population has to rely on translation models and subti
750.
by GaggiX 2y ago
Bitcoin has never reached the proliferation of AI, even though some have used it to pay for pizzas. It's 2024, and with AI you can already do: semantic search, moderation, translation, captioning, TTS, STT, context-aware grammar checke

More ›